COVID-19: The Heart of the Matter—Pathological Changes and a Proposed Mechanism

Abstract

The SARS-CoV-2 virus has resulted in over 88 million cases worldwide of COVID-19 as of January 2021. The heart is one of the most commonly affected organs in COVID-19, but the nature and extent of the cardiac pathology has remained controversial. It has been shown that patients infected with SARS-CoV-2 can sustain type 1 myocardial infarction in the absence of significant atherosclerotic coronary artery disease. However, many patients present with small elevations of troponin enzymes of unclear etiology which correlate with overall COVID-19 disease outcome. Early autopsy reports indicated variable levels of typical lymphocytic myocarditis, while radiology reports have indicated that myocarditis can be a persistent problem after recovery from acute illness, raising concern about participation in college athletics. In this communication, we review the literature to date regarding the gross and microscopic findings of COVID-19 cardiac involvement, present the findings from over 40 cases from our academic medical center, and propose mechanisms by which patients develop small elevations in troponin.  

Introduction

According to the World Health Organization, as of January 2021 there have been over 88 million reported cases of SARS-CoV-2 infection, and over 1.9 million deaths globally. ¹ Multiple organs have been shown to be affected by the virus, and resulting immune response, with cardiac involvement being among the most common. ^2,3 In patients with moderate to severe COVID-19, evidence has suggested that up to 30% may have involvement of the heart by viral particles, often leading to cardiac injury, ^{4 -7} while the prevalence of cardiac involvement in mild cases, or in those with so-called “long haul” COVID-19 remains uncertain. This review will briefly discuss the biology of SARS-CoV-2, and focus on the gross and microscopic effects reported in the heart with possible mechanistic proposals.

ACE2 and the Cellular Biology of the SARS-CoV-2 Virus in the Heart

The SARS-CoV-2 virus is a single-stranded RNA virus, closely related to the agent responsible for the severe acute respiratory syndrome pandemic of 2002-2003 (SARS-CoV) and the agent responsible for the Middle Eastern Respiratory Syndrome epidemic of 2012 (MERS-CoV). All 3 agents emerged from animal hosts to infect humans, and both SARS-CoV and SARS-CoV-2 bind to the angiotensin-converting enzyme 2 (ACE2) receptor on target cells through a spike glycoprotein. ⁸ SARS-CoV and SARS-CoV-2 utilize proteases, including TMPRSS2, to achieve fusion between the viral and target cell membranes after cell entry. ⁹ TMPRESS2 has been shown to coexpress with ACE2 in type II pneumocytes, ⁹ but less is known regarding distribution in other tissues that express ACE2. The affinity of SARS-CoV-2 for the ACE2 receptor has been reported to be greater than that of SARS-CoV, ^9,10 possibly resulting in relatively greater deposition of SARS-CoV-2 in those tissues expressing ACE2, which include the commonly infected type 2 pneumocytes and enterocytes. ACE2, however, is widely expressed in the human body, including renal tubules, cardiomyocytes, ductal cells, and the endothelial and smooth muscle cells of the vasculature. ^9,11 Through spike protein attachment to this receptor, the cardiovascular system may become exposed to not only direct viral damage, but subsequent targeting by the host immune response. There is considerable overlap among the comorbidities present between patients severely affected by SARS, MERS, and COVID-19, which include hypertension, diabetes, and previous cardiovascular disease. ⁵ The exact prevalence of significant viral heart infection and/or cardiac injury in the 3 infections is not clear. Two autopsy series of those who succumbed to the SARS epidemic in Toronto showed that SARS-CoV genome was present in 35% (7/20) and 40% (7/18) of autopsied decedents. ^12,13 In those studies, positivity was determined by RT-PCR of cardiac tissue and the authors reported that patients positive for SARS-CoV within the heart had worse outcomes, and died sooner, than those who tested negative. Histological analysis of the positive hearts showed increased fibrosis, inflammation, and hypertrophy but no significant difference in myocyte apoptosis or other forms of cell death. The authors concluded that the virus directly infected the myocardium, and such infection was associated with a more aggressive form of the disease. Other authors have concluded that COVID-19 affects the heart more frequently compared to MERS and SARS. ¹⁴ However, several studies have not shown direct infection of myocytes with the SARS-CoV-2 virus but rather have shown either no infection or infection of other cell types in the heart. ¹⁵ Very recent studies have shown that cultured induced human pluripotent stem cells (iPSC) can be infected with large viral exposure and cause direct damage to myofibrils. ^16,17 More data is needed to determine if SARS-CoV-2 can directly infect human myocytes in vivo.

Other reported immunological effects of COVID-19 include lymphopenia and abnormalities in monocytes and granulocytes. ^{6,18 -22} These effects were not reported as significant in SARS and may be specific to COVID-19. Many studies, including our own, have observed depletion of lymphocytes in the spleen and collapse of normal lymph node architecture. ^23,24 Specifically, depletion of B cells, as well as CD-4+ and CD-8+ T lymphocytes subsets and memory helper T cells (CD3+ CD4+ CD45RO+) has been described. ^25,26 This relative lymphopenia may help explain the lack of typical lymphocytic myocarditis that has been reported in autopsy studies of COVID-19 patients (see later section). ^15,27

Cardiac Pathology of COVID-19

Gross Findings

The gross cardiac findings reported in COVID-19 vary between reports, but rarely describe specific COVID-19-associated findings. Frequently reflecting the common underlying comorbidities of hypertension and diabetes, gross cardiac descriptions include hypertrophy (by organ weight), atherosclerotic coronary artery disease, scarring, and pericarditis. ^{10,15,27 -29} In one recent literature review study, 13 reported cases of myocardial infarction occurred among 277 reported autopsy cases. ³ We have seen acute dilatation of the right ventricle in the majority of our cases in which there was a concomitant elevation of brain natriuretic peptide (BNP), likely reflecting strain on the right ventricle. ¹⁵ The presence of significant blood clots in the ventricles is common, especially in the right ventricle. We have seen a clot in the right atrial appendage in one case as well as more subtle changes including clotting of the epicardial blood vessels; especially the cardiac veins (Figure 1A).

Figure 1.

Unique pathological findings of SARS-CoV-2 infection. (A) Gross cross-sections of a heart, showing right ventricular dilatation, and small vessel thrombi in the epicardial fat (green arrows); (B) acute epicardial necrosis subjacent to an occluded cardiac vein; (C) neutrophil aggregates within small cardiac vessels; (D) histological picture of a heart from a patient with death following COVID-19 due to MIS-A, showing a mixed inflammatory endotheliitis/vasculitis, including neutrophils; and (E) CD68 immunostain, highlighting monocytes within cardiac vessels.

Histological Findings

The histological changes in reported autopsy studies vary widely. Reported findings include macro and microthrombi involving both external and intramyocardial cardiac vessels, amyloid, lymphocytic myocarditis, and epicarditis/pericarditis. ^{2,15,28,30 -33} The reported frequencies of these findings varies considerably and may be related to the lack of uniform diagnostic criteria in COVID-19 cardiac pathology. ³ One recent study reported lymphocytic epicarditis/pericarditis in 3 of 23 autopsy cases examined. ³⁴ In a large European-based study, the frequency of pericarditis and myocarditis was reported in high numbers (13 of 19 with pericarditis, and 12 of 20 with myocarditis). ²⁸ A recent review of 277 autopsies suggested that the true incidence of myocarditis is relatively low (less than 2%) so the true overall incidence will require additional data. ³ A review of the literature shows that most cases that are diagnosed as pericarditis/myocarditis are based on clinical criteria such as EKG, echocardiography, and cardiac MRI findings. ^{35 -37} Without histological confirmation of the diagnosis, it is difficult to know the exact incidence and nature of the inflammatory cardiac changes reported with COVID-19. Similar to other studies, the majority of our decedents with death due to COVID-19 had mild degrees of troponin elevation at some point in their hospital course (17 of 38 cases; normal ≤0.04 ng/mL). We had 3 patients with substantial rises in troponin (4.2, 61.5, 37.4 ng/mL) who at autopsy were revealed to have myocardial infarctions. We did not find evidence of typical lymphocytic myocarditis, with or without myocyte necrosis, in any case examined. We have seen one case with diffuse amyloid deposition in the left and right ventricles (ATTR), and several cases with lymphocytic infiltrates in the superficial epicardial fat, the significance of which is not clear at this time.

The presence of lymphocytic myocarditis in COVID-19 patients has been controversial since the onset of the pandemic, when studies from China indicated that myocarditis was a significant cause of mortality. ^{38 -40} Other investigators, relying primarily on imaging studies (MRI) for diagnosis, have shown significant incidence of myocarditis. ^35,36,41 However, studies based on heart tissue obtained at autopsy have shown that classic myocarditis is present in relatively few cases; as low as less than 2% if stringent histological criteria are applied. ³ Although the incidence of typical myocarditis is not high, it is still possible that a different type of inflammatory change is present in COVID-19. A study of the SARS epidemic from 2003 described an increased number of macrophages in the hearts of decedents. ¹² Basso et al described a similar finding in a recent publication on the cardiac pathology of COVID-19. ²⁸ In our own cohort, we have also observed increased numbers of CD68-positive cells in the heart and in at least one case, individual myocytes were seen undergoing necrosis/apoptosis surrounded by CD68 cells but not by lymphocytes (Figure 1E). Finally, while typical myocarditis seems to be rare in COVID-19, the heart is the most common organ affected in reports of SARS-CoV-2 related adult multisystem inflammatory syndrome (MIS-A). ^42,43 Among the patients reported in a series of MIS-A cases across the United States and United Kingdom, all 16 patients demonstrated evidence of cardiac involvement, including abnormalities on EKG, elevated troponin, or evidence of ventricular dysfunction on imaging. ⁴² Our autopsy examination of a heart from a patient with MIS-A revealed a diffuse endotheliitis of small vessels composed of mixed inflammation, including neutrophils and CD4+ > CD8+ lymphocytes (Figure 1D). This inflammation extended into the epicardial fat, without obvious involvement of the main coronary arteries, and there was no evidence of a typical lymphocytic myocarditis. ⁴³

The presence of individual cell necrosis has been reported in approximately 14% of 277 autopsies reviewed in one literature review study. ³ All but 2 of the hearts in our cohort of 38 cases showed individual cell necrosis. ¹⁵ In one case, we observed epicardial necrosis associated with a large coronary vein occlusion (Figure 1B). In the other case, there was a myocardial infarction involving the posterior lateral septum and the lateral wall of the left ventricle. Examination of the affected myocardium revealed an early infarction, approximately 24 hours old. Examination of the associated vessels revealed significant atherosclerosis, but no occlusive thrombus was noted.

The presence of microthrombi in the heart has been reported in approximately 10% of cases to date, and has been widely reported in lungs from COVID-19 patients. ^44,45 It appears this finding may be unique to patients infected with coronavirus, reflecting the nature of the immune response and/or the hypercoagulability present. ^46,47 The mechanism by which small vessel thrombosis with resulting ischemia/apoptosis may occur is not clear but may reflect either direct damage to the endothelium and/or the presence of increased numbers of neutrophils and CD68 positive cells in the heart causing localized inflammation, and subsequent additional capillary endothelial damage (see discussion of Mechanisms of Injury below).

Imaging

Since the onset of the pandemic, there have been reports of cardiac findings using radiological techniques; primarily cardiac MRI (cMRI). ³⁵ During the pandemic, cMRI has been used primarily as a screening method to detect myocarditis in patients with cardiovascular presentations, for prediction of long-term consequences in patients recovering from COVID, and for participation screening for competitive athletes. ^35,36,41,48 However, in the absence of cardiac biopsy, it is difficult to make a definitive diagnosis of myocarditis. In addition, autopsy studies to date have not shown a large prevalence of myocarditis. ^2,3,49 There seems to be support for the use of cMRI for long term followup of patients with cardiac inflammation of any cause to predict long-term outcomes. ^35,50 However, there has been controversy regarding the appropriate role for cardiac MRI in the diagnosis of acute myocarditis in COVID-19 patients. More data and long-term follow-up is needed to determine the proper use of cMRI in COVID-19 patients.

Potential Mechanisms of Cardiac Injury in COVID-19

The SARS-CoV-2 virus gains entry into target cells by binding to the ACE2 receptor through a specific spike protein expressed on the surface of the virus. Therefore, cells that express more ACE2 receptors will be high probability targets for the virus. These cells include type II pneumocytes, enterocytes, vascular endothelium, and renal tubules. ⁸ Myocardial cells also express ACE2, and there has been great interest in how and to what extent the heart is affected by the virus. Initial symptoms in most COVID-19 patients are consistent with mild, flu-like illness. Some patients, however, go on to develop pneumonia that causes profound respiratory disease, and extensive damage to the alveoli resulting in an ARDS-like picture on imaging studies. ^2,27,51 We have also learned that other organs including the kidneys, liver, endocrine organs, the eye, skin, and the central nervous system can be affected by the virus, possibly as a result of direct viral binding to ACE2, or due to immune-mediated effects of the disease. ^52,53 The common symptoms associated with other organs are related to the respective underlying function and include acute kidney injury, diarrhea, vomiting, and anosmia and ageusia. The effects on the vascular system/endothelium are potentially lethal and include pulmonary embolism, acute myocardial infarction, and stroke. Hypercoagulability has been additionally been described in numerous studies as a distinct feature of severe COVID-19, with frequently reported elevation in serum D-dimers, as well as pathological evidence of blood clot formation in multiple organs. ^{27,33,39,53 -55} IgG fractions isolated from patients with COVID-19 have been shown to promote vascular thrombosis. ⁵⁶ Whether this feature of SARS-CoV-2 infection is directly related to endothelial damage, or to alternate pro-thombotic immune mechanisms, remains to be determined, though it is likely a factor contributing to the cardiac pathology seen in COVID-19.

It is clear that COVID-19 can cause cardiac injury and that overall clinical outcomes can be predicted by following serum troponin levels. ^{57 -59} Moreover, it has been shown that COVID-19 can cause type 1 myocardial infarction in some patients despite the absence of typical plaque rupture/hemorrhage. ^60,61 What is not clear is the etiology of small elevations of troponin that are present in many patients despite the lack of significant lymphocytic myocarditis. These changes should be considered separately from the potential long term cardiac effects of COVID-19. The remainder of this section will provide information and hypotheses regarding the potential mechanisms of cardiac injury associated with small elevations of troponin in acute, lethal COVID-19 disease.

The exact immunopathogenesis of COVID-19 is still being actively investigated. When SARS-CoV-2 infects the lungs and produces severe damage through a cytopathic effect and/or the production of a cytokine burst, the patients often suffer profound hypoxia. Hypoxia and vascular damage dramatically increase pressure within the pulmonary vasculature, resulting in increased pressures on the right ventricle. The increase in the right ventricular chamber diameter and the marked increase in BNP reported by several groups support this notion as an important source of cardiac stress. ^15,58 The markedly increased stress and overall work on the heart could induce individual cell necrosis, especially in patients with underlying coronary artery disease (Figure 2).

Figure 2.

Potential mechanisms of cardiac injury due to SARS-CoV-2 infection. MDS = monocyte derived macrophages.

The infection of the endothelium/pericytes by SARS-CoV-2 raises the possibility that the virus may cause individual myocyte death primarily through localized microvascular effects. In this scenario, infection of the endothelium by SARS-CoV-2 causes dysfunction and possibly direct virally-mediated endothelial damage resulting in activation of clotting pathways, formation of small microthrombi, and localized ischemic injury. Several authors have published electron micrographs showing viral-like particles in the endothelial compartment and microthrombi of the heart as well as other organs, increasing the likelihood that direct endothelial infection does occur. ^{15,30,44,62,63}

Which specific inflammatory mediators, if any, are present and/or activated in the heart is unknown. Many cytokines have been reported to be elevated, but in severely ill patients IL-1β, IL-6, IL-10 appear to be the most common. ⁶⁴ Given the high circulating levels of cytokines, it is likely that the heart would be affected, causing activation of immune system cells, which may in turn induce localized inflammation. The presence of neutrophil extracellular traps (NETs) have been described in COVID-19 patients. ^22,65,66 NETs are webs of DNA and oxidative enzymes released from neutrophils in response to an inflammatory stimulus; usually in response to bacteria, but also in response to cytokine release. Recent studies have shown that serum from COVID-19 patients appears to promote the formation of NETs in neutrophils harvested from healthy persons. ^22,56 Anecdotally, we have observed increased neutrophils in the smaller cardiac vessels in sections of the myocardium (Figure 1C).

The presence of increased numbers of CD68 positive cells has been described in the heart, causing localized inflammation and subsequent capillary endothelial damage. ²⁸ In our own studies, we have witnessed elevated numbers of monocytes and macrophages in the blood vessels and interstitium of COVID-19 hearts, but interestingly these CD68 positive cells are not typically associated with direct myocyte damage or death. Monocytes are mononuclear phagocytic cells that can be categorized through immunophenotyping into 3 main types; (1) “classical” (CD14+/CD16-) which migrate to sites of injury and/or infection and differentiate into inflammatory macrophages; (2) “nonclassical” (CD14dim/CD16+) which patrol the vasculature; and (3) “intermediate” (CD14+/CD16+) which are mostly inflammatory in phenotype. Interestingly, both the “classical” and the “non-classical” forms are able to cross the endothelial barrier into the affected tissue, raising the possibility that they could be a source of viral spread if they were to be infected by SARS-CoV-2. ⁶⁷ Furthermore, it has been shown that SARS-CoV-2, similar to MERS-CoV and SARS-CoV, can infect monocyte-derived macrophages (MDMs) through ACE2 and ACE2-independent mechanisms. COVID-19 infection induces readily detectable morphological and inflammation-related phenotypic changes in peripheral blood monocytes, the severity of which correlate with patient outcome. ^68,69

Once monocytes differentiate into activated macrophages, they produce a host of inflammatory cytokines which can produce both direct tissue injury and attract other inflammatory cells such as neutrophils, natural killer (NK), T- and B- lymphocytes to the local site. MDM infected with SARS-Co-V have been shown to generate a wide variety of proinflammatory cytokines, and it would be logical to assume that SARS-CoV-2 infected cells would produce a similar albeit non-identical cytokine response. ^{70 -72} Collectively, these results suggest a potential mechanism whereby SARS-CoV-2 infects MDMs, which then migrate into tissue, spreading the infection and triggering localized endothelial and/or tissue injury. This may in part explain the increased numbers of CD68+ cells in the myocardium and the pattern of scattered myocyte necrosis reported in COVID-19 patients (Figure 2). Alternatively, it is possible that these cells are responding to viral particles within the endothelium, or have been activated in an autoimmune fashion that leads to a similar appearance of myocyte necrosis. ^56,73

Conclusions

In summary, the cause of small increases in troponin seen in many COVID-19 cases is not entirely clear. What is becoming clear is that histologically confirmed lymphocytic myocarditis is not common in COVID-19 and is not the likely source of the troponin elevations. We hypothesize that SARS-CoV-2 infection can cause small elevations in troponin through individual cell necrosis/apoptosis in pathways outlined in Figure 2. In one path, the viral infection causes severe injury to the lungs and through either cytokine release and/or direct cytotoxicity results in diffuse alveolar damage which in turn causes severe hypoxia, vascular congestion and an increased workload for the heart. The increased strain, present in patients which commonly have underlying cardiac disease, would cause necrosis and/or apoptosis of individual myocytes. The other pathway centers on viral infection of the endothelium/pericyte. In this scenario, direct infection could cause either endothelial dysfunction or direct damage resulting in the formation of platelet/fibrin clots in local areas of endothelial injury. This pathway also permits the possibility that systemic cytokine generation, either independently or in concert with viral infection, could cause endothelial dysfunction or damage resulting in local platelet/fibrin clotting. Finally, a possible role for monocytes and/or MDMs causing localized endothelial damage and/or direct necrosis/apoptosis is possible and is the subject of our ongoing investigations.